Abstract: A single specimen, part and counterpart of a carapace, of a horseshoe crab from the Lower Cretaceous (Albian) Glen Rose Formation in north-central Texas, forms the basis for the definition of a new genus and species, Crenatolimulus paluxyensis. The discovery represents only the fifth limuline known from the Cretaceous. Its preservational style is remarkable in that the carapace exterior is faithfully replicated by a massive overgrowth of serpulid worms.
The stratigraphic range of the order Xiphosurida Latreille, 1802 spans nearly the entire Phanerozoic, the earliest record being from the early Cambrian (Størmer 1955). However, the number of taxa upon which the order is based remains small. The superfamily Limuloidea von Zittel, 1885, which includes the horseshoe crabs, ranges from the Permian to the present day. It is represented by merely five families, several of which are documented by only one or two species. Hauschke and Wilde (1991) summarized the history of the Limuloidea, noting the stratigraphic ranges and geographical distribution of all 34 species then known. To that list must now be added two additional occurrences. Hauschke et al. (2004) recorded the paleolimulid cf. Limulitella sp. from the Triassic of Madagascar, while Yunnanolimulus luopingensisZhang, Hu, Zhou, Lü and Bai, 2009, of the family Mesolimulidae Størmer, 1952, was described from the Middle Triassic of south-west China. It should be noted that Riek and Gill (1971) considered the Mesolimulidae undefined on the basis of objective characters and, thus, Yunnanolimulus is better assigned to Limulidae von Zittel, 1885. Twenty-eight of these occurrences predate the Cretaceous, and eleven are in western Eurasia (Hauschke and Wilde 1991); none is known from North America. From Cretaceous strata, only four species of Limulidae have been recorded, and several of the species are known from single specimens. Two of these are from North America. Thus, the discovery of a new horseshoe crab from the Albian of Texas represents a substantial increase in our knowledge of the group. The new specimen is also notable because it is preserved in a remarkable fashion. The specimen is represented by moulds of the carapace interior and exterior, the exterior surface being replicated exactly by a dense mass of serpulid worms.
The aim of the present note is to describe a new genus and species of limulid, to consider it in the context of other Cretaceous species and to document the manner in which it is preserved.
Locality and Stratigraphic Setting
The specimen was found as float in the bed of the Paluxy River (at 32°14′4.06″N, 97°47′8.71″W), approximately 1,025 m WNW (as measured along the centre of the river channel) from the centre of the US 46 bridge just west of Glen Rose, Texas (Text-fig. 1), near Dinosaur Valley State Park. Although collected ex situ, the large size of the serpulid mass within which the specimen is preserved supports the interpretation that it was derived from a nearby outcrop of levels with large serpulid mounds, just upstream from the collection site.
The Glen Rose Formation is the uppermost unit of the Trinity Group, which extends in a generally north-south belt through central Texas. It is predominantly a carbonate unit, but also contains silty and clayey beds. The fauna is quite variable, mostly marine (e.g. foraminifera, numerous molluscs, echinoderms and decapod crustaceans), but contains also ichnofossil taxa of terrestrial origin such as dinosaur footprints (Langston 1974; Bishop 1983; Farlow 1987; Farlow et al. 1989; Pittman 1989; Hawthorne 1990). The lobsters and crabs (Rathbun 1935; Stenzel 1945; Bishop 1983) were recorded from the southern extent of exposure of the formation, south-west of Austin, Texas. The formation has been interpreted to represent deposition in a series of shallow-shelf, lagoonal and supratidal environments (Nagle 1968).
At the site, which yielded the horseshoe crab, about 8 m of section of what is referred to as the ‘lower member’ of the Glen Rose Formation is exposed, the lowermost 4 m of which is depicted in Text-figure 2. Beds containing bivalves (Texigryphea, cf. Arctica), trace fossils (Skolithos, Thalassinoides) and algae are present 1–2 m stratigraphically below and above the bed from which the horseshoe crab was collected. Notable at a few horizons in the section are accumulations of serpulid worms. The limuline is derived from one of several larger serpulid accumulations, mounds or patch reefs, up to 20-cm thick and 2-m wide, which occur at a single horizon. Several reef-bearing outcrops of this horizon are noted in Text-figure 1. Thus, this part of the section is primarily marginal marine or lagoonal. The primary horizon, in which dinosaur footprints occur, the so-called Main Tracklayer, for which the site is well known, is about 1.5 m below the serpulid mound layer that yielded the new limuline.
Institutional abbreviations. FWMSH, Fort Worth Museum of Science and History, Fort Worth, Texas.
Diagnosis. Prosoma with ophthalmic ridges not meeting in front of cardiac lobe; genal angles well developed, directed posteriorly or posterolaterally, not laterally. Opisthosoma hexagonal, composed of a single fused plate, typically bearing variably placed movable lateral spines; axial furrows typically distinct.
Remarks. Family definitions within the Limuloidea have varied widely. Størmer (1952) recognized three families: Paleolimulidae Raymond, 1944, Mesolimulidae Størmer, 1952 and Limulidae. The same classification can be found in that volume of the Treatise on Invertebrate Paleontology, which deals with the class Merostomata Dana, 1852 (Størmer 1955). Two additional families, Heterolimulidae Vía Boada and de Villalta, 1966 and Austrolimulidae Riek, 1955, have subsequently been named, each for single species of Triassic limulids from Spain and Australia, respectively.
Later, Riek and Gill (1971) described a new genus of Cretaceous limuloid from Australia, Victalimulus. They argued that the differences between the Limulidae and MesolimulusStørmer, 1952, the type genus of the Mesolimulidae, were essentially matters of degree and eliminated that family. The three genera that had been referred to the Mesolimulidae were transferred to other families: Mesolimulus and PsammolimulusLange, 1923 to the Limulidae and LimulitellaStørmer, 1952 to the Paleolimulidae. Holland et al. (1975) adopted this scheme. Finally, Pickett (1984) introduced the family Dubbolimulidae to accommodate a new genus of Triassic limuloid from Australia, DubbolimulusPickett, 1984.
The arrangement of five families within the Limuloidea appears to be sustainable, as each is defined on discrete bases. Paleolimulids have a prosoma with ophthalmic ridges that meet in front of the cardiac lobe and an opisthosoma with an articulated segment in advance of the telson. Dubbolimulids possess a prosoma with short genal angles that are separated by a considerable distance from the lateral margins of the opisthosoma, which has smooth lateral margins with no indication of moveable spines. Heterolimulids have an ovoid, trilobitiform carapace in which the prosoma has weakly developed genal angles and ophthalmic ridges that converge posteriorly, while the opisthosoma is of nearly the same width as the prosoma and semicircular in outline. The Austrolimulidae exhibit genal angles that are wide, drawn into slender spines, and laterally directed. The opisthosoma bears two posterior segments that are clearly defined, but ankylosed; the margin is smooth without evidence of moveable spines. Thus, these families exhibit qualitative differences from the Limulidae and are herein considered valid families.
The Limulidae has been subdivided into two subfamilies, viz. the Limulinae, including the extant genus LimulusMüller, 1785 and the Tachypleinae Pocock, 1902, which comprises the extant TachypleusLeach, 1819 and CarcinoscorpiusPocock, 1902. A recent comparison of these three genera, based upon molecular patterns (Avise et al. 1994), supports this arrangement, yielding Limulus as sister to the other two genera. There are several extinct genera within the Limulinae, but none within the other subfamily.
Diagnosis. Prosoma moderately to strongly vaulted; opisthosoma with posterior prolongations and moveable spines that decrease in size posteriorly.
Stratigraphic range. Jurassic–Recent.
Genus CRENATOLIMULUS gen. nov.
Type species. Crenatolimulus paluxyensis sp. nov., by present designation.
Derivation of name. The generic name is a combination of the New Latin crenatus for crenate, or having the margin scalloped, and Limulus, the genus of extant North American horseshoe crab. The name alludes to the scalloped margin of the opisthosoma.
Diagnosis. Large, highly vaulted limuline. Prosoma with prominent, depressed posterior rim. Opisthosoma triangular, with scalloped lateral margins and crenulate flanks.
Remarks. The highly vaulted prosoma exhibiting ophthalmic ridges that do not meet anteriorly confirm placement within the Limulinae. The broad, concave posterior margin of the prosoma, extending from the ophthalmic ridge to the genal angle and defined anteriorly by a beaded rim, is not known in any of the other genera within that subfamily. Further, the configuration of the opisthosoma is unique, the near-quadrate cross-section differing from the more common flattened form of members within the family. In addition, the crenulate nature of the flanks and scalloped margin of the anterior part of the lateral margin of the opisthosoma are seen in no other genus within the Limulidae, making the definition of a new genus clearly warranted.
Type locality and horizon. Paluxy River near Dinosaur Valley State Park (Text-fig. 1); the Glen Rose Formation (upper Aptian – lower Albian; Early Cretaceous).
Derivation of name. The trivial name refers to the Paluxy River, which exposes the sediments that yielded the specimen.
Material. The holotype, and only specimen known to date, is FWMSH 95D-69 (part and counterpart; see Plate 2).
Diagnosis. As for genus.
Description. Prosoma wider (17.2 cm) than long; axial length 9.7 cm; total length to genal angle 13.7 cm; highly vaulted, height of prosoma from rim to centre of cardiac lobe, 6.4 cm. Margin of prosoma smoothly rounded anteriorly and nearly parallel-sided laterally, converging only slightly at genal angles. Posterior margin of prosoma with broad, concave rim widening from margin of cardiac lobe to ophthalmic ridge and remaining broad to genal angle; inner margin of rim finely beaded. Cardiac lobe defined by moderately deeply incised, parallel grooves that become broader and obscure anteriorly. Axis of cardiac lobe a sharp crest posteriorly, becoming obscure anteriorly as it terminates at elevated, elongated ocelli. Ophthalmic ridges well defined from beaded rim of posterior region to approximately mid-length of prosoma. Compound eyes large, at mid-length of ophthalmic ridge.
Opisthosoma broadly triangular; length 9.7 cm, width 12.4 cm. Anterior border convex, poorly preserved; lateral border scalloped anteriorly with at least five sharp projections separated by smooth re-entrants; posterior part of lateral margin apparently not scalloped. Posterior margin narrow, deeply concave; telson not preserved. Axial region of opisthosoma as wide as cardiac lobe anteriorly, narrowing slightly posteriorly to about mid-length; bounded laterally by narrow axial furrows in which are six depressed apodemes. Lateral regions of opisthosoma smooth and flat axially and with very steeply inclined flanks; flanks with two prominent ridges extending slightly posteroventrally, defining a tripartite surface.
Internal and ventral structures not reflected on carapace.
Occurrence. The holotype, collected as float in the bed of the Paluxy River near Dinosaur Valley State Park (Text-fig. 1), is derived from the Glen Rose Formation, upper Aptian – lower Albian (Early Cretaceous).
Discussion. The characters noted as diagnostic for Crenatolimulus paluxyensis gen. nov., sp. nov. readily distinguish it from all previously named Cretaceous limulines, as is shown later.
Victalimulus mcqueeniRiek and Gill, 1971 (Early Cretaceous, Victoria, Australia) bears a prosoma with three protuberances on the axial crest of the cardiac lobe, axial furrows that converge anteriorly and nearly meet axially, ophthalmic ridges that do not converge anteriorly and an opisthosoma with convex margins, a distinct free lobe; in addition, marginal spines are long and posteriorly directed.
Mesolimulus syriacus (H. Woodward, 1879) (early Late Cretaceous, Hâkel, Lebanon) exhibits a prosoma that is wider than long, with a narrow, parallel-sided cardiac lobe with scalloped margins. The opisthosoma is wider than long and has margins with five movable and six fixed spines.
Casterolimulus kletti Holland, Erickson and O’Brien, 1975 (Maastrichtian, Late Cretaceous, North Dakota) has a weakly vaulted prosoma, ophthalmic ridges that converge anteriorly, but do not reach the midline, deep axial furrows, converging anteriorly, and a cardiac lobe with a smooth axial crest. The opisthosoma is not known.
Limulus coffiniReeside and Harris, 1952 (Cretaceous, Colorado) lacks a prosoma, while the opisthosoma is vaulted and has straight sides, bearing six fixed spines and a broad, relatively deep, V-shaped posterior margin. The axial ridge is pronounced, bearing two blunt nodes. Six pits on either side of the axis extend from the anterior margin to mid-length at which point large, concave-forward muscle scars separate anterior from posterior parts of opisthosoma.
The preservational style of Crenatolimulus paluxyensis gen. nov., sp. nov. is remarkable. The specimen consists of a mould of the exterior of the cuticle and a mould of the interior of the same, but with no traces of cuticle preserved. It is presumed that the two parts of the fossil would have been separated from one another by a distance corresponding to the thickness of the cuticle (Text-fig. 3).
The mould of the interior of the cuticle (Plate 2, figs 1–3) consists of a fine grainstone matrix that is laced with vermiform traces. Traces are separated from one another in contrast with those on the mould of the exterior, described later. The organisms responsible for the formation of these traces are not known; however, it is likely that they were scavenging the remains of the horseshoe crab. During the initial study of the specimen, artificial resin casts of the interior and exterior of the cuticle were available for study. They are illustrated in Plate 1 and show absolutely no evidence of the texture of the sediment or of the traces. The surface of the mould of the interior of the cuticle is smooth; no evidence of the traces is visible on that surface. It is presumed that the mould of the interior is a replica of the interior of the cuticle.
The mould of the exterior of the cuticle (Plate 2, figs 4–5) is replicated entirely by a tangled mass of serpulid worm tubes. The tubes attain a maximum width of 1 mm on the surface of the fossil and are densely packed, covering the entire surface of the horseshoe crab. As with the mould of the interior of the cuticle, an artificial cast of the exterior surface shows no evidence of the presence of the serpulids. Instead, it is a faithful replica of the surface of the cuticle, which is now missing. Thin sections taken of the matrix adjacent to the fossil (Text-fig. 4A–C) show sections through the serpulid tubes surrounded by the carbonate grainstone matrix within which the specimen was preserved.
Crenatolimulus paluxyensis gen. nov., sp. nov. must have been invested with serpulids very soon after death. The cuticle of limulines is not calcified and would have deteriorated rapidly had it not been protected. Larvae of serpulids probably settled on the exposed carapace in vast numbers and began secreting their exoskeletons. Although incremental rates undoubtedly vary with geographical and ecological conditions, initial growth is very rapid. In aquarium studies, Behrens (1968) reported growth rates up to 2.5 mm per day for the first 3 days and up to 1.3 mm per day for the next 4 weeks. At rates approaching this, the entire exoskeleton could have been covered within days or weeks of death of the animal.
Attachment of sedentary polychaetes to the cuticle of fossil limulids has been reported previously. Hauschke et al. (2004) noted the presence of numerous spirorbids on the surface of two specimens from the Triassic of Madagascar and suggested that the horseshoe crabs had lain exposed on the surface for a long time prior to burial, that the animals were old resulting in a lengthy time between moults or that they did not have a burrowing habit. The specimens were encased within siderite concretions and apparently had cuticle preserved.
Serpulid accumulations within the Glen Rose Formation occur as mounds or patch reefs. The worms use firm substrates, such as molluscan shells or the horseshoe crab, as a base of attachment. Patch reefs of this type are typically formed in shallow subtidal (0,6–2,6 m) or intertidal depths in modern settings in Texas (Andrews 1964; Ten Hove and van den Hurk 1993). These conditions are consistent with the depositional setting of the Glen Rose Formation, which contains beds with Skolithos and Thalassinoides, as well as dinosaur trackways in a 5-m stratigraphic thickness containing the new limuline described here. The grainstone substrate, on which the mounds are developed, is consistent with the preferred settlement site for serpulids in modern settings, such as Baffin Bay, Texas (Andrews 1964).
Thus, the remarkable preservational style exhibited by Crenatolimulus paluxyensis gen. nov., sp. nov. was possible as a result of the presence of the limuline carcass in a shallow, marginal marine environment, with a coarse-grained substrate forming an ideal setting for settlement and rapid growth of serpulids. Without the replication of the form of the exterior of C. paluxyensis gen. nov., sp. nov. in a very short time, it is likely that the carapace would have deteriorated and not been preserved in such an excellent state.
It is interesting to speculate on the timing of death of the specimen and the incrustation by serpulids. Behrens (1968) observed that spawning of extant serpulids in the vicinity of Port Aransas, Texas, occurred in April. Assuming similar behaviour in Cretaceous serpulids, if the horseshoe crab died just at the time of spawning of the serpulids, sometime during the spring, it would provide a clean, firm surface of attachment for the young worms. The density of serpulids and their uniform size support the contention that settlement of the larvae occurred within a very brief interval of time. Subsequent growth of serpulids and addition of more individuals would have resulted in the formation of a mound.
Acknowledgements. The limuline was collected by John Hawkins (Glen Rose, Texas), who generously donated it to the Fort Worth Museum of Science and History where Aaron D. Pan, Curator of Science, assigned its catalogue number. Chris Gotcher (Waco, Texas) assisted in locating the serpulid mound layer from which the specimen originated. Careful reviews by N. Hauschke (Halle/Saale, Germany) and an anonymous referee, plus editing by J.W.M. Jagt (Natuurhistorisch Museum Maastricht, Maastricht, the Netherlands) substantially improved the manuscript. Our thanks to these individuals.